Google is reportedly developing an AI system known as “Project Jarvis,” designed to assist users by taking over web browsers to complete various tasks.
According to The Information, the system can perform functions such as gathering research, purchasing products, and booking flights. Sources familiar with the project revealed that Google could preview this concept as early as December.
Features of Project Jarvis
At the I/O 2024 event in May, Google showcased two examples of how agent-based experiences would be integrated with its Gemini AI.
The company is expected to reveal more about the agent capabilities powered by Gemini 2.0 in December. Project Jarvis, named after the AI from Iron Man, targets everyday users rather than enterprise solutions, focusing on automating routine web-based tasks.
The system operates within Google Chrome and automates tasks by taking frequent screenshots of the user’s computer screen. It interprets these images before executing actions such as clicking buttons or entering text.
However, Jarvis currently operates slowly, taking several seconds to process each action. This indicates that it likely relies on cloud processing rather than on-device capabilities.
Future Developments
Jarvis is expected to be powered by an advanced version of Google’s Gemini AI. Although a preview is anticipated in December, a full launch may take longer, as early testing is planned to address potential bugs.
Google’s approach reflects a broader trend among leading AI companies, including Microsoft and Apple, which are also developing similar technologies for improved user interaction with web content.
The report notes that the timeline for revealing Jarvis could change, as Google is evaluating its readiness for testing with a limited number of users to refine the system before a wider release.